Adaptive thermogenesis and reduced fat oxidative capacity may accompany weight loss, continuing in weight maintenance. The present study aimed (1) to determine whether weight-reduced and weight-loss relapsed women are at greater metabolic risk for weight gain compared with BMI-matched controls with no weight-loss history, and (2) to identify protective strategies that might attenuate weight loss-associated adaptive thermogenesis and support successful weight-loss maintenance. Four groups of women were recruited: reduced-overweight/obese (RED, n 15), controls (low-weight stable weight; LSW, n 19) BMI <27 kg/m2; relapsed-overweight/obese (REL, n 11), controls (overweight/obese stable weight; OSW, n 11) BMI >27 kg/m2. Body composition (bioelectrical impedance), 75 g oral glucose tolerance test, fasting and postprandial metabolic rate (MR) and substrate utilisation (RER) and physical activity (accelerometer (7 d)) were measured. Sociobehavioural questionnaires and 3 × 24 h diet recalls were completed. Fasting and postprandial MR, RER and total daily energy intake (TDEI) were not different between RED and REL v. controls (P > 0·05). RED consumed less carbohydrate (44·8 (sd 10·3) v. 53·4 (sd 10·0) % TDEI, P = 0·020), more protein (19·2 (sd 6·0) v. 15·6 (sd 4·2) % TDEI, P = 0·049) and increased physical activity, but behaviourally reported greater dietary restraint (P = 0·002) compared with controls. TDEI, macronutrient intake and physical activity were similar between OSW and REL. REL reported higher subjective fasting and lower postprandial ratings of prospective food consumption compared with OSW. Weight-reduced women had similar RMR (adjusted for fat-free mass) compared with controls with no weight-loss history. Increased physical activity, higher protein intake and greater lean muscle mass may have counteracted weight loss-associated metabolic compensation and highlights their importance in weight-maintenance programmes.